arXiv cs.AI· Liang Yuan, Maryam Khanian Najafabadi, Mashud Rana, Varvara Vetrova·· 3 天前AI 评分26
LLM 辅助编码安全性的系统性文献综述
Security in LLM-Assisted Coding: A Systematic Literature Review
AI 导读
一篇被 ACM Computing Surveys 接收的系统性综述指出，LLM 虽加速软件开发，却因训练数据偏差频繁生成不安全代码和错误补丁，模型自身也易受直接攻击。现有评测方法存在数据污染问题且缺乏稳健的安全指标，因此人工监督仍不可或缺。研究认为，更安全的自动化集成需要安全感知的模型架构、更强的安全专用基准以及更有效的人机协作框架。
正文 · 原文
Abstract:Large Language Models accelerate software development but introduce severe security vulnerabilities. This systematic review investigates security risks in automated coding, benchmark limitations, and the role of human developers. Findings indicate that models frequently generate insecure code and faulty patches due to training data biases. Additionally, the models themselves are susceptible to direct attacks. Current evaluation methods suffer from data contamination and lack robust security metrics. Consequently, human oversight remains essential. Safer integration of automation will require more security-aware model architectures, stronger security-specific benchmarks, and more effective human-AI collaboration frameworks to mitigate these systemic risks.
| Comments: | 35 pages Accepted for publication in ACM Computing Surveys (CSUR). A copy of the formal acceptance letter can be verified via the publisher's portal or provided to moderators upon request |
| Subjects: | Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI); Software Engineering (cs.SE) |
| Cite as: | arXiv:2610.15482 [cs.CR] |
| (or arXiv:2610.15482v1 [cs.CR] for this version) | |
| https://doi.org/10.48550/arXiv.2610.15482 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Liang Yuan [view email]
[v1]
Sun, 11 Oct 2026 09:33:01 UTC (1,396 KB)
来源：arXiv cs.AI · arxiv.org